Patterns of yolk testosterone deposition in two populations of Arctic-breeding Redpolls

The levels of maternal androgens in egg yolk vary considerably within and among clutches. In most species studied to date, high levels have been found to increase offspring competitive ability. We measured yolk testosterone in two populations of Redpolls (Carduelis flammea/hornemanni) nesting around Toolik Lake Field Station and in Barrow, Alaska to test three predictions about the patterns of yolk testosterone concentrations. First, we predicted yolk testosterone to increase with laying sequence so as to minimize the deleterious effects of hatching asynchrony on chicks from later hatched eggs in a species without delayed incubation. Second, we predicted yolk testosterone to be higher in Barrow where food is more abundant and Redpoll body condition is higher. Third, we predicted yolk testosterone to correlate positively with egg mass and yolk mass, which are two other indicators of maternal investment. An initial analysis that did not consider egg and yolk mass lent support to the first of these predictions. However, more thorough statistical analysis revealed that relationships between yolk testosterone and laying sequence and between yolk testosterone and site were not direct but rather driven by associations among yolk testosterone, egg mass, and yolk mass. While yolk testosterone concentration itself did not differ between sites, the relationship between yolk testosterone and egg and yolk mass did. Yolk testosterone correlated positively with egg mass at Barrow and showed a significant quadratic association with yolk mass in Toolik. While laying sequence and yolk testosterone appeared to be correlated when egg and yolk mass were not considered, the full statistical model suggests that egg and yolk mass are directly associated with laying sequence while yolk testosterone concentrations are not. These results suggest complicated relationships among egg mass, yolk mass, yolk testosterone, site, and laying sequence that should be taken into account in future studies. They also suggest that deposition of yolk testosterone may be a regulated rather than being a purely passive process.